The Journal of Nutrition Vol. 128 No. 8 August 1998, pp. 1376-1384

Addition of Triglycerides with Arachidonic Acid or Docosahexaenoic Acid to Infant Formula Has Tissue- and Lipid Class-Specific Effects on Fatty Acids and Hepatic Desaturase Activities in Formula-Fed Piglets^1,2

Sylvia de la Presa-Owens, Sheila M. Innis³, and and France M. Rioux⁴

Department of Paediatrics, University of British Columbia Vancouver, Vancouver, BC, Canada V5Z 4H4

	ABSTRACT

Abstract Introduction Methods Results Discussion References

The effects of including triglycerides with arachidonic [20:4(n-6)] or docosahexaenoic acid [22:6(n-3)] in formula on plasma chylomicron, LDL and HDL, liver, heart, kidney and brain (n-6) and (n-3) fatty acids were investigated in formula-fed piglets. Piglets were fed formula with (in % total fatty acids) 20% 18:2(n-6) and 2% 18:3(n-3) without or with 0.8% 20:4(n-6) or 0.3% 22:6(n-3) from birth to 18 d. The effects of adding 20:4(n-6) or 22:6(n-3) to the formula differed among different tissues and lipids, with the brain showing resistance to change. Piglets fed formula with 20:4(n-6) had significantly higher plasma, heart and kidney phospholipid and triglyceride, and liver triglyceride 20:4(n-6), but lower plasma and tissue phospholipid 18:2(n-6) than piglets fed formula without 20:4(n-6). Supplementation with 22:6(n-3), in contrast, had no effect on plasma or tissue 18:2(n-6). Higher 22:6(n-3) in liver phospholipid (30-92% greater) and triglyceride (200% greater) in piglets fed formula with 22:6(n-3) rather than without 22:6(n-3) was accompanied by lower 20:4(n-6) in liver phosphatidylethanolamine (mean ± SEM, 8.6 ± 0.4 and 10.5 ± 0.4% fatty acids, respectively), but higher 20:4(n-6) in triglyceride (5.2 ± 0.4 and 11.5 ± 0.5%, respectively), and higher liver, heart and kidney phospholipid 20:5(n-3). These results indicate competitive interaction between dietary 20:4(n-6) and tissue 18:2(n-6), and between dietary 20:4(n-6) and tissue 20:5(n-3), rather than 22:6(n-3). The results also show that even at low intakes, dietary 22:6(n-3) or 20:4(n-6) supplementation alters the tissue phospholipid 20:4(n-6) to 20:5(n-3) balance. Studies on the physiologic effects of dietary 20:4(n-6) and 22:6(n-3) supplementation should consider the different sensitivity among tissues to dietary fatty acids.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Large amounts of the long-chain polyunsaturated fatty acids arachidonic acid [20:4(n-6)] and docosahexaenoic acid [22:6(n-3)] are needed for membrane lipid synthesis during growth and development (Innis 1991). Recent attention has focused on the needs of the developing central nervous system (CNS)⁵ for (n-6) and (n-3) fatty acids because of the possibility that decreased CNS levels of 20:4(n-6) and 22:6(n-3) may have functional consequences (Bourre et al. 1989, Innis 1991 and 1997, Neuringer et al. 1984). Arachidonic acid (20:4n-6) and 22:6(n-3) can be synthesized from the dietary essential fatty acids, linoleic acid [18:2(n-6)] and -linolenic acid [18:3(n-3)], respectively, in brain and liver (Moore et al. 1991, Sprecher et al. 1995). It is not clear, however, if the CNS usually derives 20:4(n-6) and 22:6(n-3) by uptake from plasma, or by uptake then desaturation of the 18:2(n-6) and 18:3(n-3) precursors. Similarly, it is unclear which lipoproteins and lipids are involved in the transfer of (n-6) and (n-3) to the brain. Information has been published to suggest that both chylomicrons and VLDL, as well as other lipids, may be involved in the delivery of (n-3) fatty acids to the CNS (Anderson et al. 1994, Innis 1991, Scott and Bazan 1989).

Human milk fatty acids usually contain 8-16% 18:2(n-6), 0.5-1% 18:3(n-3), 0.5-0.7% 20:4(n-6), 0.2- 0.4% 22:6(n-3) as well as small amounts of other (n-6) and (n-3) fatty acids (Innis 1992). In contrast, infant formulas with vegetable oils as the source of polyunsaturated fatty acids contain 18:2(n-6) and 18:3(n-3) but not 20:4(n-6) or 22:6(n-3). As a result, blood lipid levels of 20:4(n-6) and 22:6(n-3) are lower in infants fed these formulas than in breast-fed infants or infants fed formula supplemented with 20:4(n-6) and 22:6(n-3) (Auestad et al. 1997, Carlson et al. 1992a and 1993, Innis et al. 1996, Makrides et al. 1995). Although recent studies have shown that infants can convert 18:2(n-6) to 20:4(n-6) and 18:3(n-3) to 22:6 (n-3) (Carnielli et al. 1996, Demmelair et al. 1995), it is not clear if the rates of synthesis are adequate to meet the needs of growing tissues. Some studies have suggested that development of visual function and scores on some other tests of neurodevelopment may be lower in term gestation infants fed formula without 22:6(n-3) than in breast-fed infants or infants fed formula with 22:6(n-3) (reviewed in Innis 1997). Other studies, however, have not found differences in visual or other tests of CNS development among infants fed milk or formulas with and without 22:6(n-3). The reason for the discrepancies among the findings of different studies with term gestation infants is not certain. Thus, the potential role of dietary 20:4(n-6) and 22:6(n-3) in facilitating neurodevelopment of young infants is currently an important area of study.

Previous studies in piglets have shown that diet-induced differences in blood lipid 20:4(n-6) and 22:6(n-3) are accompanied by similar changes in the same fatty acid in liver, kidney and other organ phospholipids, although not necessarily in brain (Rioux et al. 1997, Wall et al. 1994). Although information on the physiologic importance of reduced 20:4(n-6) and 22:6(n-3) levels in developing tissues other than the CNS is limited, it seems reasonable to consider ways to include oils with 20:4(n-6) and 22:6(n-3) in the diet of infants who cannot be breast-fed. Some studies, however, have shown that feeding formula with fish oils to provide 22:6(n-3) may reduce blood lipid levels of 20:4(n-6) and growth in young infants (Carlson et al. 1992a, 1992b and 1996). It is not clear if these effects were due to 20:5(n-3) rather than 22:6(n-3), the total amount of (n-3) fatty acid added or some other component of the fish oil, or if the decreased 20:4(n-6) and growth were causally related. The potential for adverse effects on (n-6) fatty acid metabolism due to feeding oils with 22:6(n-3) has led to the suggestion that formula with 22:6(n-3) should also contain 20:4(n-6) (ESPGAN 1991, Huang and Schmidt 1996). However, little is known to date about the effects of feeding oils enriched in 20:4(n-6) during growth and development.

The studies in this report investigated the effect of adding triglycerides with 20:4(n-6) or 22:6(n-3) to formula on tissue phospholipid (n-6) and (n-3) fatty acids of rapidly growing formula-fed piglets. In vitro studies have suggested that the (n-6) and (n-3) fatty acids compete for a common series of desaturase enzymes and that the metabolites of 18:2(n-6) and 18:3(n-3) inhibit desaturation (Innis 1991). Thus an investigation of whether addition of 20:4(n-6) or 22:6(n-3) to formula inhibits the desaturation of 18:2(n-6) or 18:3(n-3) by liver microsomes was included. With the exception of fish oil, most dietary 20:4(n-6) and 22:6(n-3) are probably in the form of animal tissue phospholipids. A further objective, then, was to elucidate the pathways of transport of 20:4(n-6) and 22:6(n-3), as chylomicron phospholipid or triglyceride, when provided in the diet as triglycerides.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Animals and diets.  Newborn male piglets weighing >1 kg at birth and <12 h old (Peter Hill Holdings, Langley, Canada) were randomly assigned to be fed one of three formulas (n = 6/group). A reference group of piglets (n = 6) remained with their mother and was fed sow's milk. Piglets in a diet group were not from the same litter. The piglets fed sow's milk are considered a reference group because of the many components in milk, e.g., growth factors, that are absent from formula and that could influence lipid and lipoprotein metabolism, as well as growth in the milk-fed animal.

The formula-fed piglets were bottle-fed by hand (Innis and Dyer 1997) with formula containing (% total fatty acids) 20% 18:2(n-6) and 2% 18:3(n-3) and no 20:4(n-6) or 22:6(n-3), or the same formula with 0.3% 22:6(n-3) added as a high 22:6(n-3), low 20:5(n-3) fish (tuna) oil, [20:5(n-3) to 22:6(n-3) ratio of 1:4] or 0.8% 20:4(n-6) from a 20:4(n-6)-rich single-cell triglyceride oil (Table 1). The amounts of 22:6(n-3) and 20:4(n-6) added were chosen to approximate the amounts of 20:4(n-6) and 22:6(n-3) in human and sow's milk (Innis 1992, Table 1). The macro- and micronutrient composition of the formula has been published (Innis and Dyer 1997). The formula and the triglycerides with 22:6(n-3) or 20:4(n-6) were provided by Ross Laboratories, Columbus, OH. The procedures involving the piglets were approved by the Animal Care Committee of the University of British Columbia and conformed to the guidelines of the Canadian Council on Animal Care.

Sample collection and analyses.  The piglets were anesthetized (ketamine/rompun, 37.5:3.75 mg/kg, MTC Pharmaceuticals, Cambridge, Canada; Bayvet Division, Chenango, Etobicoke, Canada, respectively, by intramuscular injection) at 18 d of age, 3-4 h after the last feed; blood samples were drawn by cardiac puncture (Innis and Dyer 1997). The animals were killed by intracardiac injection of 1 mol/L KCl, and the brain, liver, heart and kidney were removed, weighed and homogenized (5 mL/g, 0.32 mol/L sucrose, 15 mmol/L Tris HCl with 1 mmol/L EDTA, 1 mmol/L Mg/Cl and 1.5 mmol/L glutathione, pH 7.4). Aliquots of the liver homogenates were taken for preparation of microsomes (Nakamura et al. 1994); the remaining homogenates were stored frozen at 80°C until analysis.

Triglyceride-rich chylomicrons were isolated from fresh plasma at density 1.006 kg/L after a single ultracentrifugation at 141,000 × g for 60 min. LDL and HDL (density 1.063-1.21 kg/L) were then separated and recovered (Redgrave et al. 1975) by further ultracentrifugation, 141,000 × g for 66 h at 15°C. The small band corresponding to the VLDL of the plasma collected 3-4 h after feeding was not analyzed. The absence of apolipoprotein (apo) A and apo B in the LDL and HDL fractions, respectively, was confirmed by SDS-polyacrylamide gel electrophoresis (Maguire et al. 1989).

Tissue and lipoprotein lipids were extracted and the phospholipids, cholesterol esters and triglycerides separated on TLC plates (Whatman PK6F Silica Gel 60 A) with petroleum ether/diethyl ether/acetic acid (85:15:3 v/v/v) as the solvent system. The bands were visualized with 2',7'-dichlorofluorescein (Supelco, Bellafonte, PA), and the triglyceride and phospholipid fractions recovered. The phospholipids were then separated by using two-dimensional TLC with chloroform/methanol/acetic acid/water (100:60:16:8 v/v/v/v) and chloroform/methanol/acetic acid/water (50:30:0.5:3 v/v/v/v) in the first and second dimension, respectively. Fatty acid components were converted to their respective methyl esters, separated, identified and quantified by gas liquid chromatography (GLC).

Assay of desaturase activity.  Desaturase activities were determined in vitro using fresh liver microsomes by assay of the conversion of [1¹⁴C]-labeled 18:2(n-6) and 18:3(n-3) (American Radiolabelled Chemicals, St. Louis, MO) to their more highly unsaturated homologues. Substrate and cofactor reaction mixtures (Nakamura et al. 1994 and Purvis et al. 1983, respectively) were prepared immediately before use. Protein was determined according to Lowry et al. (1951). Desaturase enzyme reaction products were converted to their respective methyl esters, then separated based on unsaturation (i.e., saturates and fatty acids with 1, 2, 3, 4, 5 or 6 double bonds) on silver nitrate-impregnated plates with fatty acid methyl esters prepared from egg total lipid as unlabeled carriers (Innis and Yuen 1988). The plates were developed by using toluene, followed by toluene/acetone (95:5 v/v) in the same dimension, with a lane containing unlabeled authentic fatty acid methyl esters as standards on each TLC plate. Bands corresponding to the substrate [dienes or trienes, for 18:2(n-6) and 18:3(n-3), respectively], and their more highly unsaturated products, as well as the silica from the remaining area of the plate, were recovered and radioactivity quantitated using a Beckman Model Liquid Scintillation Spectrophotometer (Beckman Instruments, Palo Alto, CA). The specific activity of the 18:2(n-6) and 18:3(n-3) substrate was not corrected for the amount of the respective fatty acid in the microsomal phospholipid. The unsaturation of the fatty acid methyl esters separated by unsaturation was confirmed by GLC.

View larger version (53K): [in this window] [in a new window]   Fig 1. Levels of 18:2(n-6), 20:4(n-6), 20:5(n-3) and 22:6(n-3) in plasma chylomicron, LDL and HDL phospholipids of piglets fed as follows: solid bars, formula without 20:4(n-6) or 22:6(n-3); open diagonal stripe, formula with 20:4(n-6); shaded diagonal stripe, formula with 22:6(n-3); solid line, sow's milk (n = 6/group). Values for formula-fed piglets shown in bars are the means + SEM; values for piglets fed sow's milk are the mean (solid line) ± SEM (dotted lines). Small SEM values may not show in some plots. Note that the scale differs for each fatty acid. *Significantly different from reference group fed sow's milk; asignificantly different from group fed formula without 20:4(n-6) or 22:6(n-3) (solid bar).

Statistical analysis.  The data for piglets fed formula were analyzed using one-way ANOVA followed by preplanned comparison of the effects of addition of 20:4(n-6) or 22:6(n-3) to formula (Table 2). The effects of feeding each formula compared with the reference group of piglets fed sow's milk were determined similarly using one-way ANOVA. When the F-test indicated a significant effect, differences were analyzed by using protected least significant differences with the  level set at 0.05. All of the statistical procedures were performed using the SAS statistical software routine PROC GLM (SAS Institute, Cary, NC). Values are means ± SEM, n = 6.

	RESULTS

Abstract Introduction Methods Results Discussion References

Growth.  The addition of 20:4(n-6) or 22:6(n-3) to the formula had no significant effect on body weight, weight gain relative to formula intake, or the liver, heart or kidney weight of piglets fed the formula. Piglets fed the formula with 22:6(n-3), however, had a significantly higher brain weight, but not brain/body weight, than piglets fed the unsupplemented formula or piglets fed sow's milk (P<0.05) The liver to body weight ratio of piglets fed the formula with 22:6(n-3), but not of piglets fed the formula with 20:4(n-6), was significantly lower than that of piglets fed the unsupplemented formula.

Plasma lipoprotein fatty acids.  As expected, piglets fed the formula with 0.8% fatty acids as 20:4(n-6) had significantly higher levels of 20:4(n-6) in their chylomicron, LDL and HDL phospholipids (Fig. 1) and triglycerides (Fig. 2) than piglets fed the formula without 20:4(n-6). Similarly, addition of 0.3% 22:6(n-3) to the formula resulted in significantly higher 22:6(n-3) in the chylomicron, LDL and HDL phospholipids (Fig. 1) and in the LDL and HDL, but not chylomicron triglycerides (Fig. 2) of the formula-fed piglets. Piglets fed the formula with 22:6(n-3) also had significantly higher 20:5(n-3) in chylomicron, LDL and HDL phospholipids, and LDL, but not chylomicron or HDL triglycerides than piglets fed the formula without 22:6(n-3) (Table 3).

View larger version (39K): [in this window] [in a new window]   Fig 2. Levels of 18:2(n-6), 20:4(n-6), 20:5(n-3) and 22:6(n-3) in plasma chylomicron, LDL and HDL triglycerides of piglets fed as follows: solid bars, formula without 20:4(n-6) or 22:6(n-3); open diagonal stripe, formula with 20:4(n-6); shaded diagonal stripe, formula with 22:6(n-3); solid line, sow's milk (n = 6/group). Values for formula-fed piglets shown in bars are the means + SEM; values for piglets fed sow's milk are indicated as the mean (solid line) ± SEM (dotted line). Small SEM values may not show in some plots. Note that the scale differs for each fatty acid. *Significantly different from reference group fed sow's milk; asignificantly different from group fed formula without 20:4(n-6) or 22:6(n-3) (solid bar).

Of note, the addition of 20:4(n-6) to the formula resulted in significantly lower levels of 22:6(n-3) in chylomicron triglycerides of the formula-fed piglets. Similarly, piglets fed the formula with 22:6(n-3) had significantly lower 20:4(n-6) in chylomicron triglycerides than piglets fed the formula without 22:6(n-3). In contrast to the chylomicron, the levels of 22:6(n-3) or 20:4(n-6) were not lower in the LDL and HDL triglycerides or phospholipids of piglets fed the formula with rather than without 20:4(n-6) or 22:6(n-3), respectively. The addition of 20:4(n-6) to the formula resulted in a consistently lower level of 18:2(n-6) in chylomicron, LDL and HDL phospholipids, although not triglycerides, of the formula-fed piglets. In contrast, the addition of 22:6(n-3) had no significant effect on the levels of 18:2(n-6) in the lipoprotein lipids of the formula-fed piglets.

Piglets fed the formula with (% fatty acids) ~20% 18:2(n-6) and 2% 18:3(n-3), and no 20:4(n-6) or 22:6(n-3) had significantly higher levels of 18:2(n-6) in HDL phospholipids (Fig. 1) and in chylomicron, LDL and HDL triglycerides (Fig. 2) than the reference groups of piglets fed sow's milk. The levels of 20:4(n-6), 20:5(n-3) and 22:6(n-3) in the LDL triglycerides, but not the levels of 20:4(n-6) and 22:6(n-3) in chylomicron, LDL and HDL phospholipids were significantly lower in piglets fed the unsupplemented formula than in the milk-fed piglets. Piglets fed the formula with 0.8% 20:4(n-6) had significantly higher levels of 20:4(n-6) in chylomicron, LDL and HDL phospholipids and triglycerides, and those fed the formula with 0.3% 22:6(n-3) had significantly higher 22:6(n-3) in chylomicron, LDL and HDL phospholipids, and LDL and HDL triglycerides than the piglets fed sow's milk. These results are probably explained by the higher 20:4(n-6) and 22:6(n-3) in the formula than in the milk [0.4% 20:4(n-6), 0.2% 22:6(n-3)] (Table 1). Of note, the addition of 22:6(n-3) to the formula increased the LDL and HDL phospholipid, but not triglyceride levels of 20:5(n-3) to values significantly above those of the group fed sow's milk.

Liver fatty acids.  Again, as might be expected, the liver triglyceride levels of 20:4(n-6) were significantly higher in piglets fed the formula with 20:4(n-6) than in piglets fed the formula without 20:4(n-6) (Fig. 3). The liver phospholipids were less responsive than the triglycerides to increases in 20:4(n-6) by dietary 20:4(n-6). Although levels 20:4(n-6) were consistently higher in all of the liver phospholipids of piglets fed the formula with rather than without 20:4(n-6) (Fig. 3), the differences between the groups were not siginifcant (P > 0.05). Piglets fed the formula with 22:6(n-3), however, had significantly higher levels of 22:6(n-3) in their liver triglycerides and phospholipids than piglets fed the formula without 22:6(n-3). The levels of 22:6(n-3) in the supplemented formula (0.3% fatty acids) were lower than those for 20:4(n-6) (0.8% formula fatty acids), suggesting that liver lipid levels of 22:6(n-3) are not as tightly regulated as those of 20:4(n-6). The inclusion of 20:4(n-6) in the formula, however, resulted in significantly lower levels of 20:5(n-3), but had no effect on the levels of 22:6(n-3) in the liver phospholipids of the formula-fed piglets. Piglets fed the formula with 22:6(n-3), on the other hand, had significantly higher levels of 20:5(n-3) in liver triglycerides, phosphatidylethanolamine and phosphatidylinositol than piglets fed the formula without 22:6(n-3). Of note, although the level of 20:4(n-6) in liver phosphatidylethanolamine was significantly lower in piglets fed the formula with rather than without 22:6(n-3) (mean ± SEM, n = 6, 24.1 ±0.6 and 20.4 ± 0.4%, respectively), the level of 20:4(n-6) was greater in the liver triglycerides (9.5 ± 0.5 and 5.2 ± 0.4%, respectively). As noted in the plasma phospholipids, the addition of 20:4(n-6) but not 22:6(n-3) to the formula resulted in significantly lower 18:2(n-6) in the liver phosphatidycholine, phosphatidylethanolamine and phosphatidylserine of the formula-fed piglets.

View larger version (43K): [in this window] [in a new window]   Fig 3. Levels of 18:2(n-6), 20:4(n-6), 20:5(n-3) and 22:6(n-3) in liver triglycerides (TG), phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS) and phosphatidylinositol (PI) of piglets fed as follows: solid bars, formula without 20:4(n-6) or 22:6(n-3); open diagonal stripe, formula with 20:4(n-6); shaded diagonal stripe, formula with 22:6(n-3); solid line, sow's milk (n = 6/group). Values for formula-fed piglets shown in bars are the means + SEM; values for piglets fed sow's milk are indicated as the mean (solid line) ± SEM (dotted line). Small SEM values may not show in some plots. Note that the scale differs for each fatty acid. *Significantly different from reference group fed sow's milk; asignificantly different from group fed formula without 20:4(n-6) or 22:6(n-3) (solid bar).

The liver triglyceride and phospholipid levels of 20:4(n-6) and 22:6(n-3) were not significantly different between piglets fed the unsupplemented formula and piglets fed sow's milk, with the exception that 22:6(n-3) was higher in the liver phosphatidylcholine of the formula-fed piglets. Piglets fed the formula with 20:4(n-6) had significantly higher levels of 20:4(n-6) in liver triglycerides, phosphatidylcholine, phosphatidylethanolamine and phosphatidylinositol than the sow's milk group. Similarly, piglets fed the formula with 22:6(n-3) had significantly higher levels of 22:6(n-3) in all of the liver lipids except phosphatidylserine, and higher 20:5(n-3) in liver triglycerides and phosphatidylethanolamine than piglets fed sow's milk (Fig. 3). Levels of 18:2(n-6), on the other hand, were significantly higher in all of the liver lipids, except phosphatidylinositol, of the formula-fed piglets than of those fed sow's milk. Of note, liver phospholipid levels of 18:2(n-6) were reduced in the group fed the formula with 20:4(n-6) to values not different from those of piglets fed sow's milk.

Heart fatty acids.  As found for the liver, piglets fed the formula with 20:4(n-6) had significantly higher levels of 20:4(n-6) in heart triglycerides, phosphatidylcholine, phosphatidylethanolamine and phosphatidylinositol, and lower 18:2(n-6) in phospholipids than piglets fed the formula without 20:4(n-6) (Fig. 4). Similarly, piglets fed the formula with 22:6(n-3) had significantly higher 22:6(n-3) in heart triglycerides and phosphatidylethanolamine, and higher 20:5(n-3) in heart triglycerides, phosphatidylcholine, phosphatidylethanolamine and phosphatidylinositol than piglets fed the formula without 22:6(n-3) (Fig. 4). The addition of 20:4(n-6) to the formula had no significant effect on the heart lipid levels of 22:6(n-3), although levels of 20:5(n-3) in heart phosphatidylethanolamine and phosphatidylinositol were significantly lower in piglets fed the formula with rather than without 20:4(n-6). The addition of 22:6(n-3) to the formula had no significant effect on the levels of 20:4(n-6) in the heart lipids; again, unlike the effect of addition of 20:4(n-6), addition of 22:6(n-3) to the formula had no effect on heart 18:2(n-6).

View larger version (40K): [in this window] [in a new window]   Fig 4. Levels of 18:2(n-6), 20:4(n-6), 20:5(n-3) and 22:6(n-3) in heart triglycerides (TG), phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS) and phosphatidylinositol (PI) of piglets fed as follows: solid bars, formula without 20:4(n-6) or 22:6(n-3); open diagonal stripe, formula with 20:4(n-6); shaded diagonal stripe, formula with 22:6(n-3); solid line, sow's milk (n = 6/group). Values for formula-fed piglets shown in bars are the means + SEM; values for piglets fed sow's milk are indicated as the mean (solid line) ± SEM (dotted line). Small SEM values may not show in some plots. Note that the scale differs for each fatty acid. *Significantly different from reference group fed sow's milk; asignificantly different from group fed formula without 20:4(n-6) or 22:6(n-3) (solid bar).

Piglets fed the unsupplemented formula had significantly lower levels of 20:4(n-6) in heart phosphatidylethanolamine and higher 18:2(n-6) in all heart lipids than piglets fed sow's milk. Piglets fed the formula with 20:4(n-6) had significantly higher levels of 20:4(n-6) in the heart triglycerides and phosphatidylcholine, and piglets fed the formula with 22:6(n-3) had significantly higher 22:6(n-3) in heart triglycerides, and higher 22:6(n-3) and 20:5(n-3) in heart phospholipids than those fed sow's milk. As found in the liver, addition of 20:4(n-6) but not addition of 22:6(n-3) to the formula decreased 18:2(n-6) in the heart phospholipids to levels not different from those of the reference group fed sow's milk.

Kidney fatty acids.  Piglets fed the formula with 20:4(n-6) had significantly higher levels of 20:4(n-6) in kidney triglycerides, phosphatidylcholine, phosphatidylethanolamine and phosphatidylinositol than piglets fed the formula without 20:4(n-6) (Fig. 5). Similarly, levels of 22:6(n-3) were significantly higher in kidney triglycerides and phosphatidylethanolamine of piglets fed the formula with rather than without 22:6(n-3). As found for the heart, addition of 20:4(n-6) to the formula had no significant effect on levels of 22:6(n-3), but levels of 20:5(n-3) were significantly decreased in the kidney phosphatidylethanolamine and phosphatidylinositol of the formula-fed piglets. Feeding the formula with 22:6(n-3), in contrast, resulted in significantly higher levels of 20:5(n-3) in all of the kidney phospholipids of the formula-fed piglets. Again, as found for the heart, addition of 22:6(n-3) to the formula had no significant effect on levels of 20:4(n-6) or 18:2(n-6) in the kidney lipids. Addition of 20:4(n-6), in contrast, led to significantly lower 18:2(n-6) in the heart phospholipids of the formula-fed piglets.

View larger version (42K): [in this window] [in a new window]   Fig 5. Levels of 18:2(n-6), 20:4(n-6), 20:5(n-3) and 22:6(n-3) in kidney triglycerides (TG), phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS) and phosphatidylinositol (PI) of piglets fed as follows: solid bars, formula without 20:4(n-6) or 22:6(n-3); open diagonal stripe, formula with 20:4(n-6); shaded diagonal stripe, formula with 22:6(n-3); solid line, sow's milk (n = 6/group). Values for formula-fed piglets shown in bars are the means + SEM; values for piglets fed sow's milk are indicated as the mean (solid line) ± SEM (dotted line). Small SEM values may not show in some plots. Note that the scale differs for each fatty acid. *Significantly different from reference group fed sow's milk; asignificantly different from group fed formula without 20:4(n-6) or 22:6(n-3) (solid bar).

Piglets fed the unsupplemented formula had significantly higher levels of 18:2(n-6) in all of the kidney lipids, lower 20:4(n-6) in phosphatidylethanolamine and lower 22:6(n-3) in phosphatidylinositol than the piglets fed sow's milk. The addition of 20:4(n-6) to the formula increased 20:4(n-6) in kidney triglycerides and phosphatidylinositol, and addition of 22:6(n-3) increased 22:6(n-3) in triglycerides and phosphatidylethanolamine to levels significantly higher than those in the reference group fed sow's milk. As in the liver and heart, the addition of 20:4(n-6) to the formula decreased the kidney phospholipid levels of 18:2(n-6) of the formula-fed piglets to values not different from those of piglets fed sow's milk. Piglets fed the formula with 22:6(n-3), on the other hand, had significantly higher kidney lipid 18:2(n-6) and 20:5(n-3) than piglets fed sow's milk (Fig. 6).

View larger version (34K): [in this window] [in a new window]   Fig 6. Levels of 18:2(n-6), 20:4(n-6), 20:5(n-3) and 22:6(n-3) in brain phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylinositol (PI) of piglets fed as follows: solid bars, formula without 20:4(n-6) or 22:6(n-3); open diagonal stripe, formula with 20:4(n-6); shaded diagonal stripe, formula with 22:6(n-3); solid line, sow's milk (n = 6/group). Values for formula-fed piglets shown in bars are the means + SEM; values for piglets fed sow's milk are indicated as the mean (solid line) ± SEM (dotted line). Small SEM values may not show in some plots. Note that the scale differs for each fatty acid. *Significantly different from reference group fed sow's milk; asignificantly different from group fed formula without 20:4(n-6) or 22:6(n-3) (solid bar).

Brain fatty acids.  The addition of 20:4(n-6) or 22:6(n-3) to the formula had no significant effect on the levels of 20:4(n-6), 22:6(n-3) or 20:5(n-3) in the brain phospholipids, with the exception of a significantly lower 20:4(n-6) in brain phospatidylethanolamine of piglets fed the formula with rather than without 22:6(n-3). Piglets fed the formula with 20:4(n-6) had significantly lower levels of 18:2(n-6) in brain phosphatidylcholine and phosphatidylethanolamine than piglets fed the formula without 20:4(n-6). Consistent with the results for liver, heart and kidney, no evidence of an inverse relation between dietary 22:6(n-3) and tissue 18:2(n-6) was found in brain. Thus, piglets fed the unsupplemented formula and piglets fed the formula with 22:6(n-3) had significantly higher brain phosphatidylcholine levels of 18:2(n-6) than piglets fed sow's milk. Brain phosphatidylcholine levels of 18:2(n-6) in piglets fed the formula with 20:4(n-6), on the other hand, were not different from those of piglets fed sow's milk. Brain phosphatidylcholine levels of 22:6(n-3) in piglets fed the formula with 20:4(n-6), but not in piglets fed the formula without 20:4(n-6), were also significantly lower than in the piglets fed sow's milk.

Desaturase enzyme activities.  The in vitro hepatic microsomal desaturase enzyme assays with 18:2(n-6) or18:3(n-3) as the substrates for the 6 desaturase quantitated synthesis of fatty acid products with 3 or 4 double bonds, and for 5 desaturase, synthesis of fatty acids with 4 or 5 double bonds, respectively. In these reactions, the products of 6 desaturation serve as substrate for 5 desaturase. The results show that the addition of small amounts of 20:4(n-6) to formula (0.8% fatty acids, representing <0.5% daily energy intake) had no apparent significant inhibitory effect on the 6 desaturation of 18:2(n-6) or 18:3(n-3) by isolated liver microsomes from formula-fed piglets. Indeed, significantly higher 5 desaturation products of both 18:2(n-6) [including 20:4(n-6) and 22:4(n-6)] and 18:3(n-3) [including (20:5n-3 and 22:5(n-3)] were formed by liver microsomes of piglets fed the formula with 20:4(n-6) compared with piglets fed the formula without 20:4(n-6). In contrast, hepatic microsomes from piglets fed the formula with 22:6(n-3) showed lower rates of formation of 6 but not 5 desaturation products of 18:3(n-3) compared with piglets fed the formula without 22:6(n-3).

The rates of hepatic microsomal desaturation of 18:2(n-6) were significantly higher in all groups of formula-fed piglets than in the reference group fed sow's milk. Rates of desaturation of 18:3(n-3), on the other hand, were not significantly different between piglets fed the formula without 22:6(n-3) and piglets fed sow's milk. Piglets fed the formula with 22:6(n-3), on the other hand, had a significantly lower rate of 6 but not of 5 desaturation of 18:3(n-3) than the piglets fed sow's milk. The recovery of pentaene products [e.g., 20:5(n-3), 22:5(n-3)] from the desaturation of 18:3(n-3) was not greater in piglets fed the formula with 22:6(n-3); this lower recovery of 6 desaturase product than that in piglets fed the unsupplemented formula or sow's milk suggests inhibition of 6 desaturase activity by long-chain (n-3) fatty acids, rather than increased utilization of the products as substrates for 5 desaturation.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The study described here clearly shows that addition of triglycerides containing 22:6(n-3) or 20:4(n-6) to formula is efficacious in increasing plasma and tissue levels of 22:6(n-3) and 20:4(n-6), respectively, in formula-fed piglets. Some studies have reported lower growth in premature infants fed formula containing fish oil to provide 22:6(n-3) than in infants fed similar formula without fish oil (Carlson et al. 1992b and 1996). In addition, lower cerebrum weights were found in piglets fed formula with 4% 18:3(n-3) rather than 1% 18:3(n-3) (Arbuckle et al. 1994), or sow milk with 1.5% compared with 0.1% 22:6(n-3) (Arbuckle and Innis 1993). Recent studies have also noted lower growth in term gestation infants fed formula with 16% 18:2(n-6) and 4% 18:3(n-3) rather than 1% 18:3(n-3) (Jensen et al. 1997). In contrast, the studies reported here found no evidence of decreased growth in rapidly growing piglets fed formula with 2% 18:3(n-3) and 0.3% 22:6(n-3) from a high 22:6(n-3), low 20:5(n-3) fish oil. The formula with 22:6(n-3) had ~0.15% dietary energy as 22:6(n-3) and <0.04% energy as 20:5(n-3). Whether the finding of adverse effects of dietary (n-3) fatty acids on growth is related to the total n-3 fatty acid intake, the (n-6)/(n-3) fatty acid balance of the formula, or other components or effects of the supplemented oils is not known. The study with formula-fed term piglets reported here suggests that supplementation of formula with (% fatty acids) ~20% 18:2(n-6) and 2.0% 18:3(n-3) with up to 0.3% 22:6(n-3) from an oil low in 20:5(n-3) is unlikely to influence growth. Previous studies, however, have found that the composition of saturated fatty acids and the amount of 18:2(n-6) in the formula can influence the effect of dietary (n-3) fatty acids on tissue and plasma levels of 20:4(n-6) and 22:6(n-3) (Innis et al. 1996, Wall et al. 1994). Thus, caution is required in extrapolating the results of the studies reported here to diets of different fat content or composition, as well as to other species.

Of potential importance, these studies found lower liver triglyceride concentrations in piglets fed formula supplemented with 22:6(n-3) or 20:4(n-6) than in piglets fed an unsupplemented formula (81.3 ± 11.3, 106.1 ± 9.0 and 155.8 ± 10.1 mmol triglyceride/g protein, respectively, P < 0.05). The liver/body weight, but not total liver weight, was also lower (P < 0.05) in piglets fed the formula with 22:6(n-3) than in piglets fed the formula with 20:4(n-6) or sow's milk. The decrease in liver triglyceride associated with dietary 22:6(n-3) and 20:4(n-6) could be explained by inhibition of hepatic lipogenesis or increased fatty acid oxidation (Clark and Jump 1993, Willumsen et al. 1993, Wong et al. 1985, Wong and Nestel 1987), or decreased hepatic fatty acid uptake and re-esterification. Rates of hepatic lipogenesis, however, are low in newborn pigs (Pegorier et al. 1983); combining this with the feeding of formula providing ~50% energy from long-chain fatty acids can reasonably be expected to result in low rates of de novo fatty acid synthesis. The lower hepatic triglyceride concentrations in piglets fed formula supplemented with 20:4(n-6) or 22:6(n-3) in these studies, therefore, are more likely to be explained by changes in fatty acid oxidation, or uptake and re-esterification. Decreased plasma free fatty acid concentrations have been reported after supplementation with fish oil-derived (n-3) fatty acids (Dagnelie et al. 1994), but the mechanism of this effect does not seem to be well understood.

These studies also found lower plasma triglyceride concentrations in piglets fed formula supplemented with 22:6(n-3) than in piglets fed unsupplemented formula or formula with 20:4(n-6) (mean ± SEM, n = 6) 0.182 ± 0.040, 0.318 ± 0.004 and 0.320 ± 0.059 mmol/L, respectively). Another recent study noted no significant changes in plasma triglyceride concentrations after daily supplementation of 10 healthy adults with 1.7 g (~0.55% dietary energy) 20:4(n-6) (Nelson et al. 1997). Innis and Hansen (1996), on the other hand, found a dose-dependent decrease in plasma triglycerides in normolipidemic adults given up to 4.0 g 20:4(n-6) plus 3.0 g 22:6(n-3) per day. The lower plasma triglyceride concentrations associated with 20:4(n-6) and 22:6(n-3) supplementation in the latter study (Innis and Hansen 1996), and the lower plasma triglycerides in piglets fed the formula with 22:6(n-3) in this study are consistent with the inhibitory effect of long-chain (n-3) fatty acids on hepatic triglyceride synthesis and VLDL secretion (Wong and Nestel 1987).

The results of this study show that 22:6(n-3) increased to a greater extent in chylomicron phospholipids [from (mean ± SEM) 3.0 ± 0.2 to 5.4 ± 0.9% fatty acids] than triglycerides (0.8 ± 0.2 to 1.0 ± 0.0%) in piglets fed the formula with rather than without 22:6(n-3), suggesting that 22:6(n-3) is preferentially incorporated into phospholipids by enterocytes. Addition of triglycerides with 20:4(n-6) to the formula also resulted in a greater increase in 20:4(n-6) in chylomicron phospholipids (from 11.5 ± 0.4 to 14.9 ± 1.1%) than triglycerides (from 1.6 ± 0.3 to 3.0 ± 0.2%). In contrast, studies with rats have demonstrated higher incorporation of 20:4(n-6) when given as a free fatty acid into lymph triglycerides than phospholipids (Pavero et al. 1992). Similarly, levels of 22:6(n-3) were twofold higher in lymph triglycerides than phospholipids of rats (~350 g in body weight) given 100 or 200 mg fish oil/h by intraduodenal infusion (Clark and She 1995). In contrast to the results of the latter studies with rats, but similar to the results of these studies with piglets, inclusion of ~0.2% 22:6(n-3) in formula has been shown to increase plasma phospholipid levels of 22:6(n-3) in formula-fed infants (Innis et al. 1996). Similarly, adults given a test meal with fish oil showed a relatively higher increase in plasma phospholipid than triglyceride or cholesteryl ester (n-3) fatty acids (Nordoy et al. 1991), and plasma phospholipid (but not triglyceride) 20:4(n-6) and 22:6(n-3) were significantly increased in adults given 0.8 g 20:4(n-6) and 0.6 g 22:6(n-3)/d (Innis and Hansen 1996). In studies by Nelson et al. (1997), plasma phospholipid 20:4(n-6) increased from 10.3 to 15% fatty acids, but only from~2 to 3% in triglyceride in adults given ~ 0.55% dietary energy 20:4(n-6) per day. Possibly, at relatively low intakes, dietary 20:4(n-6) and 22:6(n-3), even when fed as triglycerides, are preferentially incorporated into phospholipids, with incorporation into triglycerides becoming important at higher intakes, possibly as the capacity for acylation into phospholipids is exceeded. Species differences, or differences in the amount or type of (n-6) and (n-3) fatty acid supplement could also explain discrepancies among the findings of studies in humans and piglets with those in rats.

In contrast to the chylomicron, the LDL and HDL, and liver triglyceride levels of 22:6(n-3) were substantially increased by including 22:6(n-3) in the formula fed to the piglets in these studies. Secretion of hepatic lipoproteins enriched in 22:6(n-3) for transport to other organs has been suggested by others (Scott and Bazan 1989). Higher levels of 22:6(n-3) in liver, LDL and HDL than chylomicron triglycerides could also be explained by relatively slow removal of 22:6(n-3) from plasma by lipoprotein lipase or slow oxidation of 22:6(n-3) in liver.

The finding in this study of lower 20:4(n-6) and 22:6(n-3) in chylomicron triglycerides of piglets fed the formula with 22:6(n-3) or 20:4(n-6), respectively, than in piglets fed the unsupplemented formula suggests competition between (n-6) and (n-3) fatty acids in the enterocyte. The lower 20:4(n-6) in liver phosphatidylethanolamine, but higher 20:4(n-6) in liver triglycerides of piglets fed the formula with 22:6(n-3) is consistent with the suggestion of Garg et al. (1989) that the decrease in liver phospholipid 20:4(n-6) after ingestion of dietary fish oil may be explained in part by a shift of 20:4(n-6) from phospholipids to triglycerides and/or cholesteryl esters. Consistent with the changes in hepatic fatty acid composition, the in vitro desaturase enzyme assays described in these studies showed no evidence of inhibition of 18:2(n-6) desaturation in piglets fed formula supplemented with 22:6(n-3). Similarly, Sprecher et al. (1994) found that addition of 20:5(n-3) to the diet, but not 18:2(n-6), 18:3(n-3) or 22:6(n-3), markedly depressed 6 desaturase activity.

The results of these studies show a greater decrease in 20:5(n-3) than in 22:6(n-3) in tissue phospholipids of piglets fed formula supplemented with 20:4(n-6). For example, levels of 20:5(n-3) in liver phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol and phosphatidylserine were decreased, but levels of 22:6(n-3) were not different between piglets fed the formula with rather than without 20:4(n-6). A decrease in 20:5(n-3) rather than 22:6(n-3) after dietary 20:4(n-6) supplementation could be explained by preferential formation of 22:6(n-3) from 18:3(n-3) or alternatively, conservation (recycling) of tissue 22:6(n-3) during inhibition of 22:6(n-3) formation.

The results of these studies with formula-fed piglets clearly show that the effects of dietary 20:4(n-6) and 22:6(n-3) differ among different tissues and lipids. For example, heart and kidney phosphatidylethanolamine, but not phosphatidylcholine or phosphatidylinositol levels of 20:4(n-6) were lower in piglets fed the formula without 20:4(n-6) than in piglets fed sow's milk. In contrast, liver phosphatidylethanolamine levels of 20:4(n-6) were not lower in the formula-fed than milk-fed piglets. Similarly, the effects of dietary essential fatty acid deficiency on 20:4(n-6) depletion differ among different tissues and phospholipids (Lefkowith et al. 1985). Unlike liver, heart and kidney appear to be unable to form 20:4(n-6) from 18:2(n-6) (Hagve and Sprecher 1989, Suneja et al. 1991). This suggests that the effects of dietary (n-6) and (n-3) fatty acid supplementation (or changes in plasma fatty acid composition) may differ between organs without desaturase enzyme activity from those in organs such as liver and brain with desaturase enzyme activity.

The results of this study also provide clear evidence for an inverse relationship between dietary 20:4(n-6) and tissue 18:2(n-6). The increase in plasma chylomicron, LDL and HDL, and liver, kidney and heart phospholipid 20:4(n-6) in piglets fed the formula with 20:4(n-6) was largely at the expense of 18:2(n-6), even though the amount of 18:2(n-6) fed was not altered. Evidence for a reciprocal relation between 18:2(n-6) and 20:4(n-6) has been reported by others (Huang et al. 1996, Innis and Hansen 1996, Nelson et al. 1997, Whelen et al. 1992 and 1993). Addition of 22:6(n-3) to the piglet formula, in contrast, did not lower the plasma or tissue phospholipid levels of 18:2(n-6) of the formula-fed piglets. Thus, the interaction between dietary 20:4(n-6) and tissue 18:2(n-6) appears to be an important aspect of essential fatty acid metabolism, separate from the competition between (n-6) and (n-3) fatty acids. As with the formula fed to piglets in these studies, many infant formulas contain higher levels of 18:2(n-6) than human milk, and blood lipid levels of 18:2(n-6) are higher in formula-fed than breast-fed infants (Innis et al. 1994 and 1996). This suggests that future studies concerning dietary (n-6) fatty acid requirements and the composition of plasma or tissue fatty acids should consider the amount of 18:2(n-6) as well as 20:4(n-6) in the diet.

In conclusion, the results of these studies have shown that feeding triglycerides containing 20:4(n-6) or 22:6(n-3) to exclusively formula-fed piglets results in tissue- and lipid class-specific changes in the (n-6) and (n-3) fatty acids of plasma chylomicron, LDL and HDL, and in liver compared with heart and kidney. The brain was relatively resistant to alteration when 20:4(n-6) and 22:6(n-3) supplementation was provided in the range of 0.15-0.4% of dietary energy, in a diet apparently adequate in 18:2(n-6) and 18:3(n-3). Whether similar differences in the sensitivity of different organs and different lipid classes occur in the human infant is not known. However, the brain weight and brain/body weight ratio of the piglet are lower, but the rate of growth is faster in piglets than in human infants. This could be interpreted such that the human infant brain is more sensitive (due to the relatively greater brain size) and organs such as liver, heart and kidney are less sensitive to dietary 20:4(n-6) and 22:6(n-3) than in piglets. Alternatively, the rapid body growth of piglets might exacerbate any limitations in dietary (n-6) and (n-3) fatty acids for the developing brain. Although caution is clearly warranted in extrapolating the results of these studies with piglets to humans, the results here suggest that future studies to explore the physiologic importance of dietary 20:4(n-6) and/or 22:6(n-3) supplementation might consider the effects on organs such as kidney and heart.

	FOOTNOTES

Manuscript received 29 September 1997. Initial reviews completed 24 November 1997. Revision accepted 17 April 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Anderson G. J., Tso P. S., Connor W. E. Incorporation of chylomicron fatty acids into the developing brain. J. Clin. Invest. 1994; 93:2764-2767
• Arbuckle L. D., Innis S. M. Docosahexaenoic acid is transferred through maternal diet to milk and to tissue of natural milk-fed piglets. J. Nutr. 1993; 123:1668-1675
• Arbuckle L. D., MacKinnon M. J., Innis S. M. Formula 18:2(n-6) and 18:3(n-3) content and ratio influence long-chain polyunsaturated fatty acids in developing piglet liver and central nervous tissue. J. Nutr. 1994; 124:289-298
• Auestad N., Montalto M. D., Hall R. T., Fitzgerald K. M., Wheeler R. E., Connor W. E., Neuringer M., Connor, S. L. Taylor, S. L., Hartmann E. E. Visual acuity, erthrocyte fatty acid composition and growth in term infants fed formulas with long chain polyunsaturated fatty acids for one year. Pediatr. Res. 1997; 41:1-10[Medline]
• Bourre J. M., Francois M., Youyou A., Dumont O., Piciotti M., Pascal G., Durand G. The effects of dietary -linolenic acid on the composition of nerve membranes, enzymatic activity, amplitude of electrophysiological parameters, and resistance to poisons and performance of learning tasks in rats. J. Nutr. 1989; 119:1880-1892
• Carlson S. E., Cook R. J., Rhodes P. G., Peeples J. M., Werkman S. H., Tolley E. A. Long-term feeding formulas high in linoleic acid and marine oil to very low birth weight infants: phospholipid fatty acids. Pediatr. Res. 1992a; 30:404-412[Medline]
• Carlson S. E., Cook R. J., Werkman S. H., Tolley E. A. First year growth of preterm infants fed standard compared to marine oil n-3 supplemented formula. Lipids 1992b; 27:901-907[Medline]
• Carlson S. E., Werkman S. H., Rhodes P. G., Tolley E. A. Visual acuity development in healthy preterm infants: effect of marine-oil supplementation. Am. J. Clin. Nutr. 1993; 58:35-42[Abstract/Free Full Text]
• Carlson S. E., Werkman S. H., Tolley E. A. . Effect of long-chain fatty acid supplementation on visual acuity and growth of preterm infants with and without bronchopulmonary dysplasia. Am. J. Clin. Nutr. 1996; 63:687-697[Abstract/Free Full Text]
• Carnielli V. P., Wattimena D.J.L., Luijendijk I.H.T., Boerlage A., Degenhart H. J., Sauer P.J.J. The very low birthweight premature infant is capable of synthesizing arachidonic and docosahexaenoic acids from linoleic and linolenic acids. Pediatr. Res. 1996; 40:169-174[Medline]
• Clark, S. D. & Jump, D. B. (1993) Regulation of hepatic gene expression by dietary fats: a unique role for polyunsaturated fatty acids. In: Nutrition and Gene Expression (Berdanier, C. D. & Hargrove, J. L., eds.), pp. 247-268. CRC Press, Boca Raton, FL.
• Clark R. M., She L. Lymphatic fatty acids from rats fed human milk and formula supplemented with fish oil. Lipids 1995; 30:673-676[Medline]
• Dagnelie P. C., Rietreid T., Swart G. R., Stijnen T., van den Berg J.W.O. Effect of dietary fish oil on blood levels of free fatty acids and ketone bodies and triacylglycerol in humans. Lipids 1994; 29:41-45[Medline]
• Demmelmair H., Schenck U. V., Behrendt E., Sauerwald T., Koletzko B. Estimation of arachidonic acid synthesis in full term neonates using natural variation of ¹³C content. J. Pediatr. Gastroenterol. Nutr. 1995; 21:31-36[Medline]
• ESPGAN Committee on Nutrition Comment on the composition of lipids in infant formula. Acta Paediatr. Scand. 1991; 80:887-896[Medline]
• Garg, M. L., Wierzbicki, A. A., Thompson, B. R. & Clandinin, M. T. (1989) W-3 fatty acids increase the arachidonic acid content of liver cholesterol and plasma triacylglycerol fractions in the rat. Biochem. J. 261:11-15.
• Hagve T. A., Sprecher H. Metabolism of long-chain polyunsaturated fatty acids in isolated cardiac myocytes. Biochem. Biophys. Acta 1989; 1001:338-344[Medline]
• Huang M. C., Craig-Schmidt M. C. Arachidonate and docosahexaenoate added to infant formula influence fatty acid composition and subsequent eicosanoid production in neonatal pigs. J. Nutr. 1996; 126:2199-2208
• Innis S. M. Essential fatty acids in growth and development. Prog. Lipid Res. 1991; 30:39-103[Medline]
• Innis S. M. Human milk and formula fatty acids. J. Pediatr. 1992; 120:S56-S61[Medline]
• Innis, S. M. (1997) Polyunsaturated fatty acid nutrition in infants born at term. In: Developing Brain and Behaviour: The Role of Lipids in Infant Formula (Dobbing, J., ed.), pp. 103-137. Academic Press, London, UK.
• Innis S. M., Auestad N., Siegman J. S. Blood lipid docosahexaenoic and arachidonic acid in term gestation infants fed formulas with high docosahexaenoic acid, low eicosapentaenoic fish oil. Lipids 1996; 31:617-625[Medline]
• Innis S. M., Dyer R. D. Dietary triacylglyerols with palmitic acid (16:0) in the 2-position increase 16:0 in the 2-position of plasma and chylomicron triacylglycerols, but reduce phospholipid arachidonic and docosahexaneoic acid and alter cholesteryl ester metabolism in formula-fed piglets. J. Nutr. 1997; 127:1311-1319[Abstract/Free Full Text]
• Innis S. M., Hansen J. W. Plasma fatty acid responses, metabolic effects and safety of microalgal and fungal oils rich in arachidonic and docosahexaenoic acids in healthy adults. Am. J. Clin. Nutr. 1996; 64:159-167[Abstract/Free Full Text]
• Innis S. M., Nelson C. M., Rioux M. F., King D. J. Development of visual acuity in relation to plasma and erythrocyte -6 and -3 fatty acids in healthy term gestation infants. Am. J. Clin. Nutr. 1994; 60:347-352[Abstract/Free Full Text]
• Innis S. M., Yuen D. E. Microsomal desaturation-elongation of linoleic acid following parenteral feeding with lipid emulsions in the rat. Lipids 1988; 23:546-550[Medline]
• Jensen C. L., Prager T. C., Fraley J. K., Chen H., Anderson R. E., Heird W. C. Effect of dietary linoleic/alpha-linolenic acid ratio on growth and visual function of term infants. J. Pediatr. 1997; 131:200-209[Medline]
• Lefkowith J. B., Flippo V., Sprecher H., Needleman P. Paradoxical conservation of cardiac and renal arachidonate content in essential fatty acid deficiency. J. Biol. Chem. 1985; 260:15736-15744[Abstract/Free Full Text]
• Lowry O. H., Roseburgh N. J., Farr A. L., Randall R. J. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 1951; 193:265-275[Free Full Text]
• Maguire G. F., Lee M., Connelly P. W. Sodium dodecyl sulfate-glycerol polyacrylamide gel slab electrophoresis for the resolution of apolipoproteins. J. Lipid Res. 1989; 30:757-761[Abstract]
• Makrides M., Neumann M, Simmer K., Pater J., Gibson R. Are long-chain polyunsaturated fatty acids essential nutrients in infancy? Lancet 1995; 345:1463-1468[Medline]
• Moore S. A., Yoder E., Murphy S., Dutton G. R., Spector A. A. Astrocytes, not neurons, produce docosahexaenoic acid and arachidonic acid. J. Neurochem. 1991; 56:518-524[Medline]
• Nakumara M. T., Tang A. B., Villanueva J., Halsted C. H., Phinney S. D. Selective reduction of  6 and  5 desaturase activities, but not  9 desaturase in micro-pigs chronically fed ethanol. J. Clin. Invest. 1994; 93:450-454
• Neuringer M., Connor W. E., Van Petten C., Barstad L. Dietary omega-3 fatty acid deficiency and visual loss in infant rhesus monkeys. J. Clin. Invest 1984; 73:272-276
• Nelson G. J., Schmidt P. C., Bartolin G., Kelley D. S., Phinney S. D., Kyle D., Silbermann S., Shaefer E. J. The effect of dietary arachidonic acid on plasma lipoprotein distributions, apoproteins, blood lipid levels, and tissue fatty acid composition in humans. Lipids 1997; 32:427-433[Medline]
• Nordoy A., Barstad L., Connor W. E., Hatcher L. Absorption of the n-3 eicosapentaenoic and docosahexanoic acids as ethyl esters and triglycerides by humans. Am. J. Clin. Nutr. 1991; 53:1185-1190[Abstract/Free Full Text]
• Pavero C., Bernard A., Cavlier H. Administration modalities and intestinal lymph absorption of arachidonic acid in rats. J. Nutr. 1992; 122:1672-1681
• Pegorier J. P., Duee P. H., Girard J., Peret J. Metabolic fate of non-esterified fatty acids in isolated hepatocytes from newborn and young pigs. Biochem. J. 1982; 212:93-97
• Purvis, J. M., Clandinin, M. T. & Hacker, R. R. (1983) Chain elongation-desaturation of linoleic acid during the development of the pig. Implications for the supply of polyenoic fatty acids to the developing brain. Comp. Biochem. Physiol. 75B: 199-204.
• Redgrave T. G., Roberts D.C.K., West C. E. Separation of plasma lipoproteins by density gradient ultracentrifugation. Anal. Biochem. 1975; 65:42-49[Medline]
• Rioux F. M., Innis S. M., Dyer R., MacKinnon M. Diet-induced changes in liver and bile but not brain fatty acids can be predicted from differences in plasma phospholipid fatty acids in formula and milk-fed piglets. J. Nutr. 1997; 127:370-377[Abstract/Free Full Text]
• Scott B. L., Bazan N. G. Membrane docosahexaenoic is supplied to the developing brain and retina by the liver. Proc. Natl. Acad. Sci. U.S.A. 1989; 86:2903-2907[Abstract/Free Full Text]
• Sprecher H., Luthria D., Geiger M., Mohammed B. S., Reinhart M. Intercellular communication in fatty acid metabolism. Fatty acids and lipids: biological aspects. World Rev. Nutr. Diet. 1994; 75:1-7[Medline]
• Sprecher H., Luthria D. L., Mohammed B. S., Baykousheva S. P. Reevaluation of the pathways for the biosynthesis of polyunsaturated fatty acids. J. Lipid Res. 1995; 36:2471-2477[Abstract]
• Suneja S. K., Nagai M. N., Cook L., Osei P., Cinti D. L. Do kidney cortex microsomes possess the enzymatic machinery to desaturate and chain elongate fatty acyl-CoA derivatives? Lipids 1991; 26:359-363[Medline]
• Wall K. M., Diersen-Schade D., Innis S. M. Plasma and tissue lipids of piglets fed formula containing saturated fatty acids from medium-chain triglycerides with or without fish oil. Am. J. Clin. Nutr. 1994; 59:1317-1324[Abstract/Free Full Text]
• Whelan J., Broughton K. S., Surette M. E., Kinsella J. E. Dietary arachidonic and linolenic acids: comparative effects on tissue lipids. Lipids 1992; 27:85-88[Medline]
• Whelan J., Surette M. E., Hardardottir I., Lu G., Golemboski K. A., Larsen E., Kinsella J. E. Dietary arachidonic acid enhances tissue arachidonate levels and eicosanoid production in Syrian hamsters. J. Nutr. 1993; 123:2174-2185
• Willumsen N., Hexeberg S., Skorve J., Lundquist M., Berge R. K. Docosahexaenoic acid shows no triglyceride-lowering effects but increases the peroxisomal fatty acid oxidation in liver of rats. J. Lipid Res. 1993; 34:13-22[Abstract]
• Wong S., Reardon M., Nestel P. Reduced triglyceride formation from long-chain polyenoic fatty acids in rat hepatocytes. Metabolism 1985; 34:900-905[Medline]
• Wong S., Nestel P. J. Eicosapentaenoic acid inhibits the secretion of triacylglycerol and of apoliprotein B and the binding of LDL in HepG2 cells. Artherosclerosis 1987; 64:139-146

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences